• 0 Posts
  • 1.3K Comments
Joined 3 years ago
cake
Cake day: June 30th, 2023

help-circle

  • The small amount of experience I have with playing around with raw hardware inputs on Linux makes me kinda surprised it took this long and guess that it was to polish this and that someone had a more or less functional version shortly after they decided to try.

    I forget the name of the system, but they have a rules system that can be set up to do arbitrary actions based on arbitrary hardware messages, without even needing to do any kind of binary driver at all.

    I used it to disable the volume commands from my soundbar while trying to get it to behave like it did with the optical input (where soundbar and PC each have their own independent volume settings), because when connected via USB, it would send the volume changes to the PC, so it looked like adjusting the volume changed it in both places. Turns out when in USB mode, it doesn’t use the soundbar volume for anything and the “double effect” was just an illusion caused by the PC steps being larger than the soundbar ones. It was nice having a system to actually check this.


  • They don’t really understand how they work and get misled by how AI can get to a correct solution (or correct looking one).

    Like I was in a meeting where people were presenting their Claude skills (which are just text files describing processes that it can add to the context) and one manager mentioned doing regression testing on added skills to make sure they don’t break the functionality of existing ones. From my pov, he was both on the right track but also missing the point entirely because they won’t be able to consistently pass regression tests even without new skills. Because something being in the context window only has a chance of affecting the output. If the code being modified has comments that look like instructions, they might override the actual instructions.

    Or it might try solving non-existent problems for you. Like a skill I was “developing” for making a particular modification to tests basically just outright said “make a test that inherits from the target test and add these parameters”. Dead simple step. First test I use to test it on, I see it’s missing one of the arguments. I mention it and the AI says that because of the start of the name being “<name of section>” and the test didn’t target that section, it decided that the argument wasn’t necessary, so I had to add instructions to not just add that argument but to not decdide to just leave it out for arbitrary reasons.

    I can’t say for sure any of the AI tasks I’ve done saved any time by being AI. But the mental load is lower and they really want us using AI, so I’ll keep doing it, but the unreliability is going to cause more problems than it solves in the long run IMO.




  • I guess the joke is that it wasn’t an ambiguous expression in the first place and that pedmas/bedmas wasn’t the issue, or rather using just it here is the problem?

    When you have multiplication expressed as numbers joined without a symbol, that takes precedence at the current layer, where layers are created using brackets, fraction symbols, superscript exponents and concatenated multiplies.

    I’m not sure this resolves all ambiguity, but it simplifies the rule to just doing multiplication/division before addition/subtraction. It seems simple enough in my mind, so I’d need to see a counter example if it does break down.

    Though I hate how mainstream math problems/puzzles always end up being an order of operations problem, which I’d argue isn’t even math but more of a metamath thing. If you’re using math to solve a real problem, the correct order of operations will be determined by logic, not any conventions.

    Like if it takes you 5 seconds to get in your car and 12 seconds per km traveled, and 5 seconds to get out of your car, if you multiply the 10 seconds to get in or out by the distance, you’ll have a wrong answer. It’ll always be distance traveled in km times 12 seconds/km plus the 10 seconds, and the math works on the units as well as the numbers to show you did it in a way that makes sense.




  • Yeah, I’ll check that out… I was thinking that primitive builder guy would do well going to the past but I wasn’t sure how much he could teach people, so it’s cool to hear about a someone doing higher tech from scratch.

    Also, that ancient puzzle box/lunar calendar/whatever it was is a counter example showing that some artisans were capable of precision work. The industrial revolution might have been more about scaling precision work to mass production levels. Like adopting standard units of measurement was a big part of it, which isn’t really technology but just getting everyone on the same page. Before that, a foot could have a different length depending on where you were, if that region even had a reliable and reproducible definition for what a foot was exactly.


  • Not fine metal, precision metal. Those ornaments didn’t have to fit something perfectly and if one person’s was slightly bigger than another’s, it didn’t really matter other than maybe for their pride.

    I’m talking about making 50 barrels with the exact same measurements (within some small error range) so that they will all fit the same receiver perfectly and can handle a standard sized bullet.

    Or, in the case of motors and machines, bearings that spin smoothly, gears that fit together without slipping, the ability to align things well enough that spinning wheels on an axle won’t add a force that wants to rip the axle apart.

    Not that electric motors are completely useless without that precision, but there’s only so far you can take them with more maintenance required without that precision.


  • Just be aware that precision metalworking wasn’t invented until the renaissance, so you might need to invent that first or your motors will wobble badly.

    Edit: that might have even been the industrial age instead of the renaissance. It might have been what really kicked off the industrial age, though the invention itself was for more reliable guns iirc.






  • I think he’s just backup for the triangle-standers there to support Grant and Sattler, but you can see all 6 of them standing in perfect form, so the poor kid probably just got bored waiting for one of them to go to the bathroom. He was supposed to have tarp ready in case of a random windstorm or helicopter, but that almost never happens so instead he was just daydreaming about giant chickens.



  • Sorry but I have figured out your secret code and will now reveal your true identity to the whole internet: You are a horse who is trying to obscure the fact that you wear horseshoes but you are also obsessed with them and can’t help but reference them, in the hopes that you’ll prove to your archnemesis, a racoon, that wearing horseshoes is way better for your hooves than the raccoon shoes he keeps trying to sell you.

    Also, be aware that he isn’t really in a horseshoes vs raccoonshoes debate with you, he’s just trying to scam you into giving him some apples for some dirty gloves he found in a dumpster. He, just like everyone else, knows very wrll that horse footwear is superior to all others and that you are hardcore af for just standing there casually while your cobbler helps you put them on with nails and a hammer.



  • One thing I’ve noticed lurking on AITA is that there’s suddenly more people casually talking about being religious. Not like overtly preaching like you’d see in the past, but more people referencing going to church or doing things for religious reasons.

    It just seemed out of place and weird. Like the tone of that part of the internet suddenly changed. It’s still more liberal than conservative, though that conflict seems to be mostly just not present, perhaps in part because of their rule against political topics, though even when some slip through, it does seem to lean more liberal or even progressive than conservative. Like plenty of abortion support, no broad support for tribal or hierarchical judgements. But it suddenly seems more religious. Christian, specifically.